Download The Dynamic Adjustment of a Transition Economy in the Early

BANK OF GREECE
THE DYNAMIC ADJUSTMENT
OF A TRANSITION ECONOMY
IN THE EARLY STAGES
OF TRANSFORMATION
Christos Papazoglou
Eric J. Pentecost
Working Paper
No. 3 May/2003
THE DYNAMIC ADJUSTMENT OF A TRANSITION ECONOMY
IN THE EARLY STAGES OF TRANSFORMATION
Christos Papazoglou
Bank of Greece, Economic Research Department
Eric J. Pentecost
Department of Economics, Loughborough University
ABSTRACT
This paper develops a model of a representative transition economy to explain the
stylised facts of output declines and real exchange rate appreciation in the early stages
of transformation. These facts can be explained by supply-side shocks, interest rate
liberalisation or a reduction in core inflation. The policy implication is that price
liberalisation in advance of financial liberalization and structural reform, including
widespread privatisation of the production process, necessarily results in some
temporary loss of output.
Keywords: Transition dynamics, overshooting, competitiveness, output decline.
JEL Classification: F41
The authors would like to thank George Tavlas and Sophocles Brissimis for helpful
comments. The views expressed in this paper are those of the authors and should in no part be
attributed to their respective institutions.
Correspondence:
Christos Papazoglou
Economic Research Department,
Bank of Greece, 21 E. Venizelos Av.,
102 50 Athens, Greece,
Tel. +30210-320.2381, Fax +30210-3233025
Email: [email protected]
1. Introduction
There are three stylized facts that characterize transition economies in the early
years of liberalization: a fall in output, a sudden sharp rise in inflation and a
depreciation of the real exchange rate followed by a slower, steady appreciation.
Table 1 shows these stylized facts for a number of Central and Eastern European
transition economies in the early 1990s. The fall in output has been attributed to
negative supply shocks (Borensztein, Demekas and Ostry 1993); a credit crunch
(Calvo and Coricelli 1992), whereby high real interest rates were imposed on
enterprises, which responded by reducing their demand for credit and production
levels; a statistical exaggeration due to under-reporting of the activity of the private
sector (Berg and Sachs, 1994); and the limited mobility of resources (Gomulka,
1998). The rise in inflation is usually attributed to the early liberalization of goods
market prices, which rise in line with world prices following administered repression,
but where output is slow to respond to the price signals, due to the slowness of the
privatization process and the lack of market-oriented institutions. For example,
according to the EBRD (1999), of their sample of 13 transition economies, there were
only two that had not liberalized the majority of goods market prices by 1992
(Romania and Ukraine), whereas only five had liberalized their financial sectors by
1995 (Czech Republic, Hungary, Poland, Slovakia and Slovenia). The depreciation
and subsequent appreciation of the real exchange rate during transition is similarly
explained by Halpern and Wyplosz (1997) as due to the initial inflation shock
followed by a gradual rise in productivity. This pattern is less clear from Table 1,
partly due to currency changes, which make consistent data difficult to obtain in the
initial transition phase, but also due to sharp changes in exchange rate policy1 and the
different speeds of transition. For countries such as the Czech Republic, Poland,
Romania and Slovakia, however, the depreciation-appreciation pattern is apparent in
Table 1.
This paper attempts to provide a theoretical rationale for these stylized facts by
setting out an aggregate model of a representative transitional economy. The model is
1
The principal transition economies initially operating a flexible exchange rate policy included:
Bulgaria, Lithuania, Moldova, Romania, Russia, Slovakia, Slovenia, Ukraine, with Hungary and
Poland having crawling pegs. However, the Czech Republic switched from a pegged rate to a managed
float in 1997, Lithuania and Bulgaria switched to a currency board in 1994 and 1997 respectively and
Poland moved to a flexible rate in 2000.
5
characterized by flexible goods prices, together with sticky money wages and
unemployment so an upward-sloping aggregate supply curve is maintained.
Moreover, the model has an underdeveloped financial sector in which the central bank
sets the short-term interest rate and there are few private capital inflows, which
although not necessarily legally restricted, are limited by fragile and thin financial
markets, which carry a high-risk premium for international investors. Domestic
residents can, however, use foreign money for domestic transactions as a means of
hedging against high inflation as well as the economic instability and uncertainty that
are characteristic of the early stages of the transition process. This approach follows
Calvo and Rodriguez (1977), Frenkel and Rodriguez (1982) and more recently, Calvo
and Frenkel (1991), in that the emphasis is on the thinness of capital markets and
currency substitution as the most appropriate features of the financial markets.
This model enables the dynamic effects of three specific ‘liberalization shocks’ to
be considered by examining the effects of once-and-for-all changes in three
exogenous variables. The ERBD (2001) shows that banking sector reform has taken
place since the mid-1990s. This is captured in the model by a rise in the nominal
interest rate, reflecting the liberalization of the financial sector as financial repression
is eased (see McKinnon, 1973). The EBRD (2001) also shows a gradual increase in
small and large-scale privatizations, which we capture by considering a rise in the real
wage rate, indicative of a more flexible supply-side. The third shock that we consider
is a reduction in core inflation, which has been a common aim of macroeconomic
policy in the transition economies since the high rates of inflation in the early 1990s
and is consistent with the inflation record of these economies shown in Table 1.
The rest of the paper is set out as follows. The model is set out in Section 2,
followed by the general solution to the model (Section 3). Section 4 examines the
dynamics of the three liberalization shocks noted above and Section 5 offers a brief
conclusion.
2. The Model
The analysis relies on a dynamic macroeconomic model, which attempts to
capture some of the basic characteristics of a typical transition economy in the early
stages of transformation. In particular, by allowing the use of foreign currency
holdings for domestic transactions - some currency substitution is permitted, while
6
interest rates are still set by the state. In the real sector, the analysis assumes the price
of domestic output to adjust much faster to any excess demand than output. This
assumption turns out to be important for the dynamics of the system. It relies on the
empirical observation that while price liberalization occurred almost immediately in
most economies, structural change proceeded gradually, reflecting the nature of such
processes, as well as limiting the flexibility of product and labor markets and thus
constraining the response of the real sector to relative price changes2. Finally, the
economy is assumed to operate below full employment.
In the goods market, real demand for domestic output depends directly upon the
level of real income, the real interest rate and the real exchange rate. That is,
z = γy − σ (r − π ) + θc
(1)
where the variables of the model are expressed in logarithmic form and
z = real aggregate demand,
y = the real volume of domestic output,
r = the short-term, nominal interest rate (not in logs), taken to be fixed,
c = p * + e − p = the relative price of foreign to domestic goods,
p = the price of domestic output,
e = the current exchange rate, measured in units of domestic currency per unit
of foreign currency,
p* = the foreign price level,
π ≡ p& = the rate of inflation of p .
The nominal, short-term interest rate is assumed to be largely determined by the
central bank, reflecting the fact that the financial system is not yet fully liberalized
and that market-forces play little role in determining nominal interest rates. It is
therefore assumed that the long and the short rate can be treated as identical due to the
underdeveloped capital markets and financial sector. Thus there is no term structure
of interest rates and the same interest rate enters the aggregate demand for goods
function as the money-demand functions.
The financial sector consists of two liquid assets - domestic money, h and foreign
money, h*. Domestic residents are assumed to hold both assets. The demand for
2
As a matter of fact the limited resource mobility to massive changes in relative prices constituted one
of the reasons for the slow output recovery according to Gomulka (1998).
7
domestic money depends positively on real income, y , and the domestic cost of
living, q , and inversely on the expected depreciation of the local currency, Ee& , and
the nominal interest rate, r . In this model the expected depreciation of the home
currency in the demand for money function reflects the fact that domestic residents
may also hold foreign currency for transactions or precautionary purposes in the
presence of domestic inflation. Thus domestic money market equilibrium is given by:
h = q + α ′y − β ′Ee& − δ ′r
(2a)
where Ee& is the expected rate of exchange depreciation and q = sp + (1 − s )( p * + e) the
domestic cost of living, where s is the share of domestic production in total demand.
The domestic demand for foreign currency (measured in terms of home currency)
is assumed to depend directly on the domestic cost of living, real income, the
expected depreciation of the home currency and the nominal interest rate that is
(e + h*) = q + α * y + β * Ee& − δ * r
(2b)
Combining (2a) and (2b), and assuming perfect foresight, such that Ee& = ε , where ε
is the actual rate of depreciation, we obtain:
(h − h * −e) = αy − βε − δr
α > 0, β > 0, δ > 0
(2)
defining α = α ′ − α * , β = β ′ + β * and δ = δ ′ − δ * . The fact that, α > 0 and δ > 0
implies that the income and interest elasticities of the demand for domestic money
exceed the corresponding elasticities for the domestic demand for foreign money3.
Equation (2) indicates that asset holders are willing to absorb a higher ratio of
domestic to foreign currency holdings, if real domestic income rises, if the expected
rate of depreciation falls or if the nominal interest rate falls. Equation (2) is the
relation for financial market equilibrium, with β indicating the degree of currency
substitution, which closely resembles the specification used by Calvo and Frenkel
(1991)4.
3
The interest rate elasticity is likely to be positive since the preferred-habitat hypothesis suggests that
domestic assets are more responsive to domestic yields than foreign assets. Note, however, that this is
primarily an empirical matter. There is very little empirical evidence appertaining to the likely values
of these elasticities in transitional economies. If industrial countries can be used as a guide, then Bordo
and Choudhri (1982) find that for Canada α lies between 0.5 and 0.7 and δ between 10.5 and 11.5.
Lahiri (1992) finds α >0 for Yugoslavia, all of which confirm the signs postulated in this paper.
4
According to Sahay and Vegh (1996), currency substitution captures the use of a foreign currency as
a means of exchange relative to domestic currency and in that sense is measured by the opportunity
cost of both currencies. Uribe (1997) and Mongardini and Mueller (2000) emphasise the importance of
ratchet effects in currency substitution. These effects are not included here for tractability.
8
On the supply-side of the economy, it is assumed that money wages are set
exogenously, but that on the demand-side labor is employed until the marginal
revenue product is equal to the money wage rate. Thus the supply of output depends
inversely on the real wage rate. That is,
y = −ν ( w − p )
(3a)
where, w = nominal wage. By adding and subtracting the cost of living, q , equation
(3a) can be expressed as,
y = −νω − τc
(3)
where, ω ≡ w − q is the real consumption wage and τ ≡ ν (1 − s ) >0. In the analysis
that follows we assume that the real consumption wage is fixed, implying that the
nominal wage is set in such a way as to keep the real wage constant.
The last static equation represents the determinants of expected inflation in p.
Expected inflation is assumed to be the same as actual inflation, π, by perfect
foresight expectations, and directly related to excess demand in the goods market plus
some exogenous rate of core inflation, π , hence
π = φ ( z − y) + π
(4)
where φ ≥ 0 denotes the speed of adjustment of inflation expectations to excess
demand in the goods market5.
The dynamic part of the model consists of three equations:
c& = π * +ε − π
(5a)
h& = µ
(5b)
h&* = η ( y − z )
(5c)
Equation (5a) describes the evolution of the relative price, c . Equation (5b) states that
the rate of growth of the domestic money supply is fixed and equal to µ . Finally,
Equation (5c) states that the rate of accumulation of foreign money reflects the extent
of excess supply in the goods market. That is, the current account surplus reflects the
increase in the foreign currency holdings of the public, i.e. the capital account deficit
(see, for example, Wilson, 1992).
By specifying the ratio of domestic to foreign money as, m ≡ h − h * −e , the
model consisting of equations (1), (2), (3), (4) and (5), can be summarized as,
5
A better specification for expected inflation should refer to the cost of living. This, however, would
complicate the analysis without altering the results in any essential way.
9
z = γy − σ (r − π ) + θc
(6a)
m = αy − βε − δr
(6b)
y = −νω − τc
(6c)
π = φ ( z − y) + π
(6d)
c& = π * +ε − π
(6e)
m& = µ − η ( y − z ) − ε
(6f)
Equation (6f) is obtained by taking the time derivative of the definition of m and then
substituting from (5b) and (5c). Turning to the evolution of the system, we assume
that at all points other than those where the exchange rate undergoes jumps in
response to unanticipated disturbances, the real exchange rate and the ratio of
domestic to foreign money evolve continuously and can be taken as predetermined.
The four equations (6a)-(6d) yield the short-run solutions for the four variables
z , y, ε , π in terms of c and m . Substituting these solutions into (6e) and (6f), the
dynamic adjustment of the system is determined.
The steady-state equilibrium, denoted by tildes, is reached by setting
c& = h& = h&* = 0 , implying
~
y=~
z
µ = ε~ = π~ = π
(7a)
(7b)
Note from equation (7b), that core inflation reflects the long-run rate of growth in the
~ and ~
money supply. More specifically, the long-run equilibrium values for c~, m
y are:
νθ
τσ
τσ
~
y =− ω +
π −
r
J
J
J
(8a)
ν (1 − γ )
σ
σ
c~ = −
ω− π + r
J
J
J
(8b)
~ = − ναθ ω −  δ + τασ r −  β − τασ π
m
J
J  
J 

(8c)
where, J = τ (1 − γ ) + θ > 0.
3. The General Solution of the System
In order to determine the evolution of the economy over time, we must first solve
the static equations (6a)-(6d) for the short-run solutions of z , y, ε and π . These
solutions are given by the following expressions:
10
z=
[θ − τ (γ − σφ )] ν (γ − σφ )
σ
σ
c−
ω− r+ π
∆
∆
∆
∆
ε =−
π =
τα
να
1
δ
c−
ω − m− r
β
β
β
β
(9a)
(9b)
φ [θ + τ (1 − γ )]
νφ (1 − γ )
σφ
1
c+
ω−
r+ π
∆
∆
∆
∆
y = −τc − νω
(9c)
(9d)
where ∆ ≡ (1 − σφ ) , which is taken to be positive. Substituting (9a)-(9d) into (6e) and
(6f), the dynamics of the economy are derived. In particular, the two roots are,
λ1 = −(φ − η ) [τ (1 − γ ) + θ ] ∆ < 0 , and λ 2 =
1
> 0 which means that the equilibrium
β
is a saddlepoint6. For λ1 < 0 , the speed of the response of product market prices is
greater than the response of the money supply to any excess demand. Note that the
gradual response of the money supply is the result of the slow adjustment of the
output market. The stable locus, which is derived using the stable root, is negatively
sloped and represented by the line SL in Figure 1.
For the system to remain stable, m and c must always be on the stable arm. Any
disturbance that causes the line SL to shift also leads to instantaneous jumps in m and
c that allow the system to move to the new equilibrium. These jumps are due to a
discontinuous adjustment in the exchange rate since prices are assumed to adjust
sluggishly to clear the goods market. According to the definitions of m and c, the
jumps in these variables are described by,
dc = de
(10a)
dm = − de
(10b)
from which it follows that, dm / dc = −1 . This equation captures the direction along
which the instantaneous jump from one stable locus to another takes place. It is
illustrated by the negatively sloped line XX in Figure 1. Note that the slope of the
stable locus, as shown in Figure 1, SL, is flatter than the XX line.
From the stable locus and the definitions in (10a) and (10b) we can derive an
equation connecting changes in the exchange rate to changes in ω , r and π , which
represent the three disturbances under consideration, that is:
6
The detailed derivations are available from the authors on request.
11
de =
1
λ2

 g 21ν (1 − γ ) ναθ
σ
φJ
δ
σ
φJ
{
}dω − { (λ1 + ) − }dr + {1 + (λ1 + )}dπ  (11)
+
J
β
J
βJ
∆
∆

 J (1 − βλ1 )
where g 21 = (τα / β ) + (η (τ (1 − γ ) + θ )) / ∆ > 0 .
Equation (11), together with the relations (10a) and (10b) and the short-run solutions
given by (9a)-(9d), form the basis for our analysis regarding the impact of the three
disturbances.
4. Disturbances
There are three disturbances to consider: a rise in the real wage, characteristic of
price liberalisation and increased privatisation of the productive sector; a rise in the
short term rate of interest, characteristic of financial sector liberalisation; and a
reduction in the core rate of inflation, brought about by a fall in the rate of monetary
growth, characteristic of a stability-oriented monetary policy7.
4.1 An increase in the real wage
A supply shock can be represented by a change in the real consumption wage, ω .
A rise in the real wage rate, relative to the marginal product of labour, will lower
competitiveness, output and the money stock. From equations (8a)-(8c), the increase
in the real wage reduces domestic output which, in turn, requires a real exchange rate
appreciation to equilibrate the product market. Furthermore, the output decline
generates an excess supply in the money market, which requires a fall in the domestic
to foreign money ratio.
The initial effect of the rise in the real wage will be to cause an immediate
depreciation of the exchange rate, since from (11), (de0 dω ) > 0 where the subscript
0 refers to the initial impact effect. This, in turn, from (10a) and (10b), implies
The effect of changing the currency substitution parameter, β , because it affects the slope of the
stable locus, may alter the impact on the model of each of the three disturbances. The results derived,
however, do not give unambiguous results, making the impact of a higher β minimal. For example,
7
consider the effect of a fall in m on c (the real exchange rate) under the assumption of a larger β . The
fall in m generates an excess demand for money, which with a higher value of β means a smaller
increase in the rate of depreciation needed to clear the money market. At the same time, the rise in
domestic inflation will also be less since output does not have to fall by as much, because the rise in the
rate of depreciation absorbs most of the excess demand in the money market. Thus these two forces
work in opposite directions making the impact of a larger β on the slope of the stable locus
qualitatively ambiguous and probably minimal.
12
dc 0
dm
de
=− 0 = 0 >0
dω
dω
dω
(12)
which means that there is an improvement in competitiveness and a fall in the
domestic to foreign money ratio. The initial jumps in c and m affect the short-run
equilibrium variables of the system z , y, ε and π . Thus, by taking the differentials of
(9a)-(9d) and using (12), we obtain
dy 0
de
= −ν − τ 0 < 0
dω
dω
(13a)
dz 0
ν (γ − σφ ) [θ − τ (γ − σφ )] de0
=−
+
∆
∆
dω
dω
(13b)
dε 0
να (1 − τα ) de0
=−
+
dω
β
β
dω
dπ 0 νφ (1 − γ ) φJ de0
=
+
>0
dω
∆
∆ dω
(13c)
(13d)
The impact of the rise in the real wage is illustrated in Figure 1. The effect of the
initial depreciation in the exchange rate is to shift the stable locus down to SL1. The
system moves from the initial steady state at point A, along the XX line to B, which
corresponds to the new short-run equilibrium, where according to (12) m is lower and
c is higher at m1 and c1 respectively. Then, from point B, c falls and m increases as the
system converges along the SL1 locus to the new steady state equilibrium at C.
In the short-run, the direct effect of an increase in the real wage is to reduce
domestic output. At the same time, the fall in the demand for domestic goods, due to
the fall in real income, leads to a real depreciation and with the price of domestic
output given instantaneously, this is achieved through an immediate depreciation of
the nominal exchange rate. This relative price effect, on the one hand, tends to offset
the impact of a lower output on aggregate demand, while on the other hand, by
enhancing the output fall, indirectly affects aggregate demand inversely. The fall in
output leads to higher inflation through the Phillips curve and this has a positive
impact on aggregate demand through a lower real interest rate. Thus, it appears that,
while in the short-run the output fall is in excess of the long-run decline, due to real
depreciation, the overall impact on aggregate demand is ambiguous. As a result, a
short-run excess demand for domestic output is generated, which corresponds to a
current account deficit. In the money market, the decline in output lowers demand
while the exchange rate depreciation, by reducing the ratio of domestic to foreign
13
money, decreases supply. This, in turn, means that the rate of exchange depreciation
may or may not increase in order to secure equilibrium in the money market.
The adjustment period is associated with rising prices and expanding domestic to
foreign money ratio, both reflecting the excess demand for domestic output. The fact
that prices respond faster than the money supply to excess demand contributes to real
appreciation of the exchange rate during adjustment. The fall in competitiveness leads
to output recovery while simultaneously reducing demand. The adjustment process
continues until the current account deficit is eliminated and the new long-run
equilibrium is achieved, with lower levels for output and competitiveness.
4.2 An increase in the interest rate
The increase in the short-term interest rate, as seen from (8a)-(8c), lowers money
demand leading to a fall in the domestic to foreign currency ratio. It also leads to a fall
in the demand for domestic output and as a result equilibrium in the goods market
requires an improvement in competitiveness.
In the short run, a rise in the interest rate leads to an immediate depreciation of
the exchange rate overshooting its long-run equilibrium. From equation (11) it follows
that (de0 dr ) > 0 where, as before, the subscript 0 refers to the initial impact effect.
Then, from (10a) and (10b) it follows that
dc 0
dm
de
= − 0 = 0 >0,
dr
dr
dr
(14)
which means that the rise in the interest rate also leads to an improvement in
competitiveness and to a decrease in the domestic to foreign money ratio.
Then, by taking the differentials of (9a)-(9d) and using (14), we deduce the
following effects on the short-run equilibrium variables of the system
dy 0
de
= −τ 0 < 0
dr
dr
(15a)
dz 0
σ [θ − τ (γ − σφ )] de0
=− +
dr
dr
∆
∆
(15b)
dε 0
δ (1 − τα ) de0
=− +
β
β
dr
dr
(15c)
dπ 0
σφ φJ de0
=−
+
.
dr
∆
∆ dr
(15d)
14
The impact of the rise in the rate of interest is also illustrated in Figure 1. The effect
of the initial depreciation in the exchange rate is to again shift the stable locus down
to position SL1. The system moves from the initial steady state at point A, along the
XX line to point B, which is the new short-run equilibrium. Then, during the
adjustment process c begins to fall and m to increase as the system moves towards the
new steady-state equilibrium at point D.
In the short-run, the rise in the interest rate reduces the demand for domestic
money and domestic output. The fall in aggregate demand causes an immediate
depreciation of the exchange rate which, given the slow adjustment in the price of
domestic output, corresponds to a real depreciation. The real depreciation, in turn,
leads to a fall in domestic output. Thus, while output falls, the impact on aggregate
demand appears ambiguous. We postulate, however, that a short-run excess demand
for domestic output is the more likely outcome leading also to higher inflation.
Turning to the financial sector, the depreciation of the exchange rate, by lowering the
domestic to foreign currency ratio, is more likely to generate an excess demand,
which means that it more than offsets the impact of a higher interest rate and lower
output. As a result, the expected depreciation of the exchange rate will increase in
order to restore equilibrium.
Both the price level and the domestic to foreign currency ratio are rising during
the adjustment process. The increase in the price of domestic output leads to lower
competitiveness, which, by stimulating output and reducing demand, starts
eliminating the current account deficit. As a result, the rate of inflation and the rate of
exchange depreciation begin falling until they become equal to the core inflation,
which reflects the rate of the exogenously given rate of monetary expansion. The
adjustment continues until the system converges to the new long-run equilibrium at
point D in Figure 1.
4.3 A decrease in core inflation
A decrease in the core rate of inflation reflects a decline in the rate of growth of
the money supply, which is representative of macroeconomic stabilisation. In the long
run, according to (8a)-(8b), the decline in the core inflation lowers demand for
domestic output, which means that an improvement in competitiveness is required to
restore equilibrium. The real depreciation, in turn, reduces domestic output. Thus, a
15
policy to reduce the rate of core inflation will again have a deleterious effect on
output, although the economy will become more competitive. In the money market,
according to equation (7b), the fall in core inflation leads to an equivalent decline in
the rate of depreciation. This increases demand while the fall in output has the
opposite effect. We postulate, however, that the former impact prevails and thus the
domestic to foreign currency ratio rises.
The decline in core inflation leads to instantaneous appreciation of the exchange
rate. Again, using equation (11), we see that (de0 dπ ) > 0 and then, from (10a) and
(10b) we obtain
dc 0
dm
de
=− 0 = 0 >0
dπ
dπ
dπ
(16)
which reflects lower competitiveness and a higher domestic to foreign money ratio.
Now, for the short-run equilibrium variables of the system, by taking the differentials
of (9a)-(9d) and using (16), we derive
dy 0
de
= −τ 0 < 0
dπ
dπ
(17a)
dz 0 σ [θ − τ (γ − σφ )] de0
= +
>0
dπ
∆
∆
dπ
(17b)
dε 0 (1 − τα ) de0
=
>0
dπ
β
dπ
(17c)
dπ 0 1 φJ de0
= +
>0 .
dπ
∆ ∆ dπ
(17d)
As illustrated in Figure 2, the initial appreciation in the exchange rate shifts the stable
locus to the right to SL1. The system moves from the initial steady state at point A,
along the XX line to the short-run equilibrium at B. Then, c starts depreciating and m
falling so that new steady state equilibrium is reached at point C.
In the short-run, a fall in core inflation, brought about by a lower rate of growth
of the money supply, leads to an immediate appreciation of the exchange rate. This
corresponds to lower competitiveness since the price of domestic output adjusts
sluggishly. The decline in relative prices leads to an increase in domestic output. The
demand for domestic output, however, falls due to the higher real interest rate and
lower competitiveness, although the increase in output works in the opposite
direction. As a result, the fall in core inflation generates an excess supply in the
domestic output market. In the money market, the exchange rate appreciation
16
increases supply, while the higher output raises demand. We postulate that the former
effect prevails and thus the rate of exchange depreciation must decrease in order to
restore equilibrium.
The excess supply in the market for domestic output brings about a fall in the
price of domestic output as well as in the domestic to foreign money ratio during the
adjustment period. The depreciation in the real exchange rate that follows removes the
current account surplus. The new long-run equilibrium at point C in Figure 2
corresponds to a situation of lower output and improved competitiveness.
5. Conclusions
In this paper we have used an aggregate dynamic macroeconomic model in order
to explain some of the stylized facts that have characterized transition economies in
the early stages of the transformation process. The analysis has incorporated some of
the basic features of the particular economies such as the potential existence of
currency substitution as well as the rapid price liberalization, as compared to the
slower pace of financial and industrial sector reforms. Thus the liberalization of
interest rates, the reduction of core inflation and the rise in real wages all serve to
reduce output in the early stages of the transition process. The reason for this initial
fall in output is an assumed lack of long-term inward investment and the slowness of
the supply-side reforms - due to the need for the creation of market-orientated
institutions - to bring about productivity gains to raise output.
Furthermore, in the cases of a rise in the real wage and financial liberalization the
adjustment period is characterised by real exchange rate appreciation. This result is in
line with the actual experience of most transition economies (Halpern and Wyplosz,
1997), and in this model is due to falling competitiveness as goods market prices
respond faster than output to excess demand in the product market. This reflects the
fact that price liberalisation is relatively easy and quick to undertake, whereas
structural reform is much more complex and effective only after a considerable time
delay. This analysis suggests that, notwithstanding sufficient structural reform, the
rapid price liberalization has a depressing influence on output in transition economies.
17
References
Berg, A., Sachs, J.D.,1992. Structural Adjustment and International Trade in Eastern
Europe: the Case of Poland. Economic Policy 14, 117-173.
Bordo, M., Choudhri, E., 1982. Currency Substitution and the Demand for Money:
Some Empirical Evidence for Canada. Journal of Money, Credit and Banking 14,
48-57.
Borensztein, E., Demekas, D.G., Ostry, J.D.1993. An Empirical Analysis of the
Output Declines in Three Eastern European Countries. IMF Staff Papers 40, 1-31.
Calvo, G., Coricelli, F. 1992. Stabilising a Previously Centrally Planned Economy,
Poland 1990. Economic Policy 14, 176-226.
Calvo, G., Frenkel, J.A.,1991. From Centrally Planned to Market Economy. IMF Staff
Papers 38, 268-299.
Calvo, G., Rodriguez, C.A., 1977. A Model of Exchange Rate Determination Under
Currency Substitution and Rational Expectations. Journal of Political Economy
85, 261-78.
European Bank for Reconstruction and Development, 1999. Transition Report: Ten
Years after Liberalisation. EBRD, London.
European Bank for Reconstruction and Development, 2001. Transition Report.
EBRD, London.
Gomulka, S., 1998. Output: Causes of the Decline and the Recovery. In Boone, P.
Gomulka, S., Layard, R. (Eds), Emerging from Communism: Lessons from
Russia, China and Eastern Europe. MIT Press, Cambridge.
Frenkel, J.A., Rodriguez, C.A., 1982. Exchange Rate Dynamics and the Overshooting
Hypothesis. IMF Staff Papers 29, 1-29.
Halpern, L., Wyplosz, C., 1997. Equilibrium Exchange Rates in Transition
Economies. IMF Staff Papers 44, 430-461.
Lahiri, A.K., 1991. Money and Inflation in Yugoslavia. IMF Staff Papers 83, 751788.
McKinnon, R.I. 1973. Money and Capital in Economic Development. Brookings
Institute, Washington.
Mongardini, J., Mueller, J. 2000. Ratchet Effects in Currency Substitution: An
Application to the Kyrgyz Republic. IMF Staff Papers 47, 218-237.
18
Sahay, R., Vegh, C., 1996. Dollarization in Transition Economies: Evidence and
Policy Implications. In Mizen, P., Pentecost, E.J. (Eds), The Macroeconomics of
International Currencies: Theory, Evidence and Policy. Edward Elgar,
Cheltenham, UK, 193-224.
Uribe, M. 1997. Hysteresis in a Simple Model of Currency Substitution. Journal of
Monetary Economics 40, 185-202.
Wilson, J.F. 1992. Physical Currency Movements and Capital Flows. In IMF Report
on the Measurement of International Capital Flows, Part II, Background
Papers, 91-97.
19
Table 1
Inflation, Output Growth and Real Exchange Rate Changes in Selected
Transition Economies 1990-96 (% per annum)
Country
Albania
Inflation
Growth
Competitiveness1
Bulgaria
Inflation
Growth
Competitiveness
Czech Republic
Inflation
Growth
Competitiveness
Hungary
Inflation
Growth
Competitiveness
Poland
Inflation
Growth
Competitiveness
Romania
Inflation
Growth
Competitiveness
Slovakia
Inflation
Growth
Competitiveness
Slovenia
Inflation
Growth
Competitiveness1
1990
1991
1992
1993
1994
1995
1996
-10
104
-28
237
-10
31
11
-24.6
16
9
-22.4
6
9
-6.6
20
5
2.9
-9
339
-12
79
-7
64
-2
53.7
122
2
-8.9
33
3
12.3
165
-4
-14.2
52
-14
-7.6
13
-6
4.6
18
-1
16.3
10
3
5.1
8
5
3.4
9
5
6.7
-4
3.7
32
-12
10.4
22
-3
8.8
21
-1
8.8
21
3
-1.0
28
2
-4.0
22
2
2.8
-12
-15.9
60
-7
56.5
44
3
6.4
38
4
7.3
29
5
1.0
22
7
8.2
19
5
8.8
-6
-32.5
223
-13
-6.9
199
-9
-38.2
296
1
38.7
62
4
7.5
28
7
-2.2
60
5
-9.6
58
-15
-3.0
9
-7
1.7
28
-4
5.5
12
5
1.0
7
7
2.8
6
6
-0.3
247
-8
93
-5
23
1
8.8
18
5
-2.6
9
4
-16.0
10
3
7.1
0
-3
-5
Sources: Inflation is end of year rate, from the EBRD Transition Report, 1999.
Growth is the growth of real output, from the EBRD Transition Report,
1999.
Competitiveness is from International Financial Statistics, Annual
Yearbook, 2001, real effective exchange rate (line rec).
Competitiveness1 is calculated from International Financial Statistics,
Annual Yearbook, 2001 using the ratio of the US consumer price index
multiplied by the average local dollar exchange rate to the local CPI,
where no series is recorded for the real effective exchange rate.
20
m
X
m0
m2
m3
m1
A
C
D
B
SL
SL1
X
c2
c0
c3
c1
c
Figure 1. Model dynamics: real wage and interest rate shocks
21
m
m1
m2
X
B
C
SL1
A
m0
SL
X
c1
c0
c2
c
Figure 2. Model dynamics: a reduction in core inflation
22